jyxent

Do you mean that this is technically possible, or that this is a feature that will be supported? I've preordered the Time 2, so having the ability to configure it that way to significantly extend the battery life of the ring could be interesting. On the other hand, using it standalone might be fine for me. I wouldn't really know how fast I might go through the battery without using it for a while.

Bluesky has also added 1m new users for each of the past 2 days.

It says that lead fumes are unlikely to be generated:

Based on standard soldering iron temperatures of 620°F-700°F and the melting point of lead (621°F), it is unlikely that lead fume will be generated during electronic soldering, unless the solder is heated to lead’s vaporization temperature of 3182°F.

I've been learning BCC / bpftrace recently to debug a memory leak issue on a customer's system, and it has been super useful.

I think XY actually has better resolution depending on your perspective. I agree that the minimum extrusion is approximately a line width circle. But that circle can move less than the line width for features on larger surfaces. For example, if you emboss something on the side of an object with 0.3mm depth using a 0.4mm nozzle / line-width, it will work fine.

Anybody who lives in a rural area with a well could have higher levels of natural occuring fluoride in their water too.

How is this Elon-hate? Is a founder someone who founded a company or someone who joined it later? Calling someone not a founder doesn't change their contributions to the company.

Not defending them, but that would explain why they thought it was 2 years old.

I'm guessing they didn't look at the article and assumed it was the older one. There is a similar one from two years ago: https://mawfig.github.io/2022/06/18/v-lang-in-2022.html

If it is happening all the time, providers of generative AI should be emphasizing a whole lot more that they can just make up information.

ChatGPT currently has this message in small text below the message prompt, but I'm not sure how long it has been there: "ChatGPT can make mistakes. Consider checking important information."

But I think that message is kind of underplaying it. The fact that it can make mistakes doesn't necessarily mean that the average person will think it is completely making something up.

And this discussion from November has an update on the progress: https://github.com/fish-shell/fish-shell/discussions/10123

And you downplay evidence of fraud. You mentioned studies where 1/3 to 1/2 of reviewed papers are found to not be replicable and say they get similar but weaker results. This is evidence of p-hacking or data manipulation. You should expect far less papers to not be replicable given proper statistics.

You misunderstand. The issue is that many more papers should be retracted. A paper this major that couldn't be replicated still took another decade to be retracted.

There are many more papers that are less cited that will never get looked at in enough detail to discover fraud.

You're missing the 'and their output.'

The licenses for the three libraries / software they mention are MIT or BSD. They don't need to make it open source.

Without watching the video referenced in the article, I'm guessing this was based on what they theoretically could do, rather than something that they had actually implemented.

They currently only use the load cells for bed leveling: https://www.youtube.com/watch?v=F7ETSVu04ao&t=940s

I could be wrong, but I would expect that the load cell readings would be far too noisy in order to determine the height of the nozzle while printing. Even if they were accurate, different materials would also behave differently due to different viscosities.

I'm not aware of any hotend sensors doing as you describe. The Prusa MK4 only uses it to generate a bed mesh the same as any other bed-leveling sensors. It does not do any live calibration while printing.

They measure a series of points to get a bed mesh, the same as other leveling sensors do. You shouldn't need to live-adjust z, because you've measured the bed surface with the nozzle tip.

With a BLTouch, you need to adjust the z-offset, because the nozzle tip and the BLTouch trigger height are not the same. Once that is calibrated, it should continue to work fine as long as the sensor is working properly.

I think it mainly comes down to the accuracy of the sensor. If your sensor gives you a +-0.03mm reading when probing the same spot, then you will have high and low spots in your bed mesh that you can't correct with z-offset.

This could explain your issues with the BLTouch you mentioned in another thread here. I have various brands of knockoff BLTouch sensors, and the worst one had a +-0.02mm range when measuring points, which made it impossible to get a good bed mesh. The one I replaced it with had about a +-0.005mm accuracy, which fixed my issues with the low and high spots.

The Bambu Lab CoreXY printers have force sensors in the bed for leveling using the nozzle: https://wiki.bambulab.com/en/x1/maintenance/replace-the-heat...

Vorons also have a nozzle probing design: https://github.com/VoronDesign/Voron-Tap

If they're using pirated books, haven't they already committed copyright infringement by keeping these copies? The model itself might be transformative, but I think that is a separate issue.

If the companies had bought all of the ebooks, then they might be in the clear.

I think that is a slightly different issue. Google and Apple charge a percentage for any purchases made on their app stores.

The equivalent here would be Prime offering a free ad-supported channel and Google requiring them to hand over a percentage of their ad revenue for the channel.

Pokemon Go had a similar peak and decline after initial launch: https://www.bbc.com/news/technology-37176782

You can restrict keys when added so that they are only available on specific hosts: https://www.openssh.com/agent-restrict.html

I don't think it is sustainable for Twitter to hand out ad revenue site-wide at this point. All they have done so far is to pay out to a small group of people. Musk has said that ad revenue is down 50% and that they are not currently profitable.

I guess it depends what endpoint you are talking about. The actual vaccine was pretty quick, it was the phases of testing that took longer: https://www.sec.gov/Archives/edgar/data/1682852/000119312520...

Adding some randomization to the exponential backoff times to avoid the thundering herd problem: https://en.wikipedia.org/wiki/Thundering_herd_problem

It never was alive. A fork where the only development is renaming everything to crablang is pointless.

Reddit reported that they had 430 million monthly active users in 2019.

$500 million per year / 430 million users = $1.16 per user per year

$1.16 / 12 months = $0.10 per user per month.

It is not entirely clear how they can change to make less requests, or if the Reddit employee even compared apps fairly. There is a floor count of API calls needed.

The Apollo developer tested the official Reddit app, and it makes a similar amount of API calls: https://www.reddit.com/r/redditdev/comments/13wsiks/api_upda...

I'm not the parent commenter, but the friction point for me is the slowness when typing the first few characters in an interactive search (I have a large history). I think the searches are synchronous with each keystroke, right? It would feel a lot faster if each keystroke could cancel an in progress search instead of waiting for it to finish.

In fzf, there is no noticeable lag when typing.